Glaucoma is the second leading cause of blindness in the world, but, with proper treatment, blindness can be avoided in 90% of individuals with glaucoma. Proper treatment begins with the detection of glaucoma. Our long-term goal is to improve the detection of early glaucomatous damage, as well as the detection of progression of such damage. In this proposal, we focus in particular on the macular region, the most important retinal region for everyday visual performance. To better understand glaucomatous damage to the macula, as part of Aim 1, we test the hypothesis that early macular visual defects have a particular, arcuate, form when tested with behavioral tests [i.e. standard automated perimetry (SAP)]. We propose an anatomical framework to understand the basis of these macular arcuate defects. Based upon this framework, specific structural (anatomical) hypotheses are generated to understand the type of patients who may be susceptible to these defects. These hypotheses are tested using SAP, multifocal visual evoked potentials, and a relatively new noninvasive technique for in vivo measurement of the anatomy of the human retina and optic nerve, called frequency domain optical coherence tomography (fdOCT). Glaucoma damages retinal ganglion cells (RGC) and their axons. Most of the in vivo anatomical studies in humans have focused on the retinal nerve fiber layer (RNFL), which is made up of axon of the RGCs. As part of Aim 2, we focus on measuring RGC thickness directly using fdOCT technology. In particular, we test a simple linear model, which relates local SAP field loss to RGC loss. In addition, we test the hypothesis that RGC loss is a more sensitive measure than peripapillary RNFL thickness for detecting macular damage Finally, in Aim 3 we use our linear structure-function model to improve our ability to detect glaucomatous damage and its progression. In particular, we use the model to predict the progression of structural and functional damage in patients with glaucoma and to predict the relative effectiveness of different tests for detecting glaucoma. Further, our theoretical framework allows us to test hypotheses about why different tests of glaucoma may or may not agree.